def haplotype_hut():
'''Load partially-phased haplotypes for the hutterites sample problem.
Cached since it's large.'''
if not Templates.HAPLOTYPE_HUT:
Templates.HAPLOTYPE_HUT = io_genotype.read('plink', 'haplotype', tped=GENOTYPE_SAMPLE + '.hap.tped',
load_ids=False)
return Templates.HAPLOTYPE_HUT
def test_nearest_snp(self):
'''Test finding the nearest SNP of a given base pair location.'''
g = io_genotype.read('plink', 'genotype', prefix=itu.GENOTYPE_SAMPLE)
assert_equal(list(g.nearest_snp_multiple(g.base_pair)), g.snp_range, 'Wrong nearest SNP location')
i = 5
assert_equal(g.nearest_snp(g.base_pair[i]+0.01), i, 'Wrong nearest SNP location')
assert_equal(g.nearest_snp(g.base_pair[i]-0.01), i, 'Wrong nearest SNP location')
def problem_hut():
'''Load the hutterites data set. Cached since it's large.'''
if not Templates.PROBLEM_HUT:
pedigree = Templates.pedigree_hut()
genotype = io_genotype.read('plink', 'genotype', prefix=GENOTYPE_SAMPLE, load_ids=False)
Templates.PROBLEM_HUT = Problem(pedigree, genotype)
return Templates.PROBLEM_HUT
def setUp(self):
'''Load test data and expected results.'''
unittest.TestCase.setUp(self)
# The way to load a pedigree in conjunction with a genotype set is to recode
# its sample IDs to consecutive for easier access by phasers.
self.problem = io.read_plink(prefix=itu.GENOTYPE_DUO, haplotype=None, pedigree=itu.GENOTYPE_DUO + '.tfam')
self.phaser = trivial_phaser()
# Expected results
self.solution = Problem(self.problem.pedigree, io_genotype.read('plink', 'genotype', prefix=itu.GENOTYPE_DUO_SOLUTION))
def test_nearest_snp(self):
'''Test finding the nearest SNP of a given base pair location.'''
g = io_genotype.read('plink', 'genotype', prefix=itu.GENOTYPE_SAMPLE)
assert_equal(list(g.nearest_snp_multiple(g.base_pair)), g.snp_range,
'Wrong nearest SNP location')
i = 5
assert_equal(g.nearest_snp(g.base_pair[i] + 0.01), i,
'Wrong nearest SNP location')
assert_equal(g.nearest_snp(g.base_pair[i] - 0.01), i,
'Wrong nearest SNP location')
def haplotype_hut():
'''Load partially-phased haplotypes for the hutterites sample problem.
Cached since it's large.'''
if not Templates.HAPLOTYPE_HUT:
Templates.HAPLOTYPE_HUT = io_genotype.read('plink',
'haplotype',
tped=GENOTYPE_SAMPLE +
'.hap.tped',
load_ids=False)
return Templates.HAPLOTYPE_HUT
def problem_hut():
'''Load the hutterites data set. Cached since it's large.'''
if not Templates.PROBLEM_HUT:
pedigree = Templates.pedigree_hut()
genotype = io_genotype.read('plink',
'genotype',
prefix=GENOTYPE_SAMPLE,
load_ids=False)
Templates.PROBLEM_HUT = Problem(pedigree, genotype)
return Templates.PROBLEM_HUT
def setUp(self):
'''Load test data and expected results.'''
unittest.TestCase.setUp(self)
# The way to load a pedigree in conjunction with a genotype set is to recode
# its sample IDs to consecutive for easier access by phasers.
self.problem = io.read_plink(prefix=itu.GENOTYPE_TRIO, haplotype=None, pedigree=itu.GENOTYPE_TRIO + '.tfam')
self.phaser = trivial_phaser()
# Expected results
self.solution = Problem(self.problem.pedigree, io_genotype.read('plink', 'genotype', prefix=itu.GENOTYPE_TRIO_SOLUTION))
def read_plink(**kwargs):
'''Load a problem from the following PLINK files:
Default Override Option Data Format
======================================================================================
prefix.pdg.tfam pedigree Pedigree adjacency PLINK TFAM
(genotyped+nongenotyped samples)
prefix.tfam pedigree_genotyped Genotyped sample pedigree
(sub-graph of the pedigree) PLINK TFAM
corresponding to prefix.tped
prefix.tped genotype Genotype data PLINK TPED
prefix.hap.tped haplotype* Haplotype data PLINK TPED
prefix.err error** Genotype errors flagged Integer array (snps x samples)
prefix.info info Problem info pickle (binary)
prefix.frm frames LD-independent SNP frames text file
prefix.lam lam*** Haplotype est. recombination rate text file
* - hap data not loaded if this option is None.
** - errors set to 0 if this file is not found or this option is set to None.
*** - data not loaded if if this file is not found.
'''
# Read input options
verbose = kwargs.get('verbose', False)
prefix = kwargs.get('prefix', '')
overrideable_option = lambda name, default: kwargs.get(name, default if prefix else None)
pedigree = overrideable_option('pedigree', prefix + '.pdg.tfam')
pedigree_genotyped = overrideable_option('pedigree_genotyped', prefix + '.tfam')
genotype = overrideable_option('genotype', prefix + '.tped')
haplotype = overrideable_option('haplotype', prefix + '.hap.tped')
error_file = overrideable_option('error', prefix + '.err')
info = overrideable_option('info', prefix + '.info')
if not np.all([[pedigree, pedigree_genotyped, genotype, error_file] is not None]):
raise ValueError('Must specify a prefix or pedigree, pedigree_genotyped, genotype, error files')
frames_file = overrideable_option('frames', prefix + '.frm')
lam_file = overrideable_option('lam', prefix + '.lam')
# Load data
print_location = lambda x : x if x else '-'
if verbose: print 'Reading pedigree from %s, %s ...' % (print_location(pedigree), print_location(pedigree_genotyped),)
p = io_pedigree.read(pedigree, genotyped_id_file=pedigree_genotyped)
if verbose: print 'Reading genotype data from %s ...' % (print_location(genotype),)
g = io_genotype.read('plink', 'genotype', tped=genotype, load_ids=False)
if verbose: print 'Reading haplotype data from %s ...' % (print_location(haplotype),)
h = io_genotype.read('plink', 'haplotype', tped=haplotype, load_ids=False) if haplotype else None
if verbose: print 'Reading error data from %s ...' % (print_location(error_file),)
error = np.loadtxt(error_file) if error_file and os.path.isfile(error_file) else None
if verbose: print 'Reading frame data from %s ...' % (print_location(frames_file),)
frames = db_gene.snp.ld_graph.read_frames(frames_file) if frames_file else None
lam = np.loadtxt(lam_file) if lam_file and os.path.isfile(lam_file) else None
# info = ProblemInfo(p, g) if info is None else info
problem = Problem(p, g, haplotype=h, error=error, frames=frames, lam=lam)
if haplotype and info:
if verbose:
print 'Reading problem info from %s ...' % (info,)
with open(info, 'rb') as fout:
problem.info = pickle.load(fout)
return problem
if options.recode:
# First, compute allele frequencies with PLINK
util.run_command('%s --nonfounders --freq --out %s' % (plink_cmd_base, out_base_name))
# Convert frequencies file that to a reference allele recoding
# file (a file containing the list of SNPs and their minor allele letter)
bu.frq_to_minor_file(out_base_name + '.frq', out_base_name + '.mnr')
# Then convert binary PLINK to a recoded 12-recoded TPED, where 1=minor allele for each SNP
out_recoded = out_base_name + '.recoded'
util.run_command('%s --transpose --recode12 --reference-allele %s.mnr --out %s' % \
(plink_cmd_base, out_base_name, out_recoded))
# Reload the recoded problem
for ext in ('nof', 'tped', 'tfam'):
os.rename(out_recoded + '.' + ext, out_base_name + '.' + ext)
genotype = io_genotype.read('plink', 'genotype', tped=out_base_name + '.tped', load_ids=False)
else:
genotype = problem.genotype
# Write problem to file in our (npz)
io.write_npz(problem, out_base_name + '.npz')
# Write genotypes Gaixin formats; she uses those separate files
io_genotype.write('gaixin', genotype, options.out_gxn + '.gxn',
sample_id=problem.pedigree.sample_id_genotyped)
# Convert plink tped to bed; delete the tped set
util.run_command('%s --make-bed --out %s' % (plink_cmd_base, out_base_name))
for ext in ('nof', 'pdg.tfam', 'tped', 'tfam', 'info'):
os.remove(out_base_name + '.' + ext)
except:
traceback.print_exc(file=sys.stdout)
def test_create_from_file_no_sample_ids(self):
g = io_genotype.read('plink',
'genotype',
tped=itu.GENOTYPE_SAMPLE + '.tped',
load_ids=False)
self.__assert_genotype_set_stats_correct(g)
def test_create_from_file(self):
g = io_genotype.read('plink', 'genotype', prefix=itu.GENOTYPE_SAMPLE)
self.__assert_genotype_set_stats_correct(g)
self.__assert_snp_equals((22, 'rs1654', 0, 17596388), g.snp[7])
def test_create_from_file_no_sample_ids(self):
g = io_genotype.read('plink', 'genotype', tped=itu.GENOTYPE_SAMPLE+'.tped', load_ids=False)
self.__assert_genotype_set_stats_correct(g)
def test_create_from_file(self):
g = io_genotype.read('plink', 'genotype', prefix=itu.GENOTYPE_SAMPLE)
self.__assert_genotype_set_stats_correct(g)
self.__assert_snp_equals((22, 'rs1654', 0, 17596388), g.snp[7])
def read_plink(**kwargs):
'''Load a problem from the following PLINK files:
Default Override Option Data Format
======================================================================================
prefix.pdg.tfam pedigree Pedigree adjacency PLINK TFAM
(genotyped+nongenotyped samples)
prefix.tfam pedigree_genotyped Genotyped sample pedigree
(sub-graph of the pedigree) PLINK TFAM
corresponding to prefix.tped
prefix.tped genotype Genotype data PLINK TPED
prefix.hap.tped haplotype* Haplotype data PLINK TPED
prefix.err error** Genotype errors flagged Integer array (snps x samples)
prefix.info info Problem info pickle (binary)
prefix.frm frames LD-independent SNP frames text file
prefix.lam lam*** Haplotype est. recombination rate text file
* - hap data not loaded if this option is None.
** - errors set to 0 if this file is not found or this option is set to None.
*** - data not loaded if if this file is not found.
'''
# Read input options
verbose = kwargs.get('verbose', False)
prefix = kwargs.get('prefix', '')
overrideable_option = lambda name, default: kwargs.get(
name, default if prefix else None)
pedigree = overrideable_option('pedigree', prefix + '.pdg.tfam')
pedigree_genotyped = overrideable_option('pedigree_genotyped',
prefix + '.tfam')
genotype = overrideable_option('genotype', prefix + '.tped')
haplotype = overrideable_option('haplotype', prefix + '.hap.tped')
error_file = overrideable_option('error', prefix + '.err')
info = overrideable_option('info', prefix + '.info')
if not np.all([[pedigree, pedigree_genotyped, genotype, error_file]
is not None]):
raise ValueError(
'Must specify a prefix or pedigree, pedigree_genotyped, genotype, error files'
)
frames_file = overrideable_option('frames', prefix + '.frm')
lam_file = overrideable_option('lam', prefix + '.lam')
# Load data
print_location = lambda x: x if x else '-'
if verbose:
print 'Reading pedigree from %s, %s ...' % (
print_location(pedigree),
print_location(pedigree_genotyped),
)
p = io_pedigree.read(pedigree, genotyped_id_file=pedigree_genotyped)
if verbose:
print 'Reading genotype data from %s ...' % (
print_location(genotype), )
g = io_genotype.read('plink', 'genotype', tped=genotype, load_ids=False)
if verbose:
print 'Reading haplotype data from %s ...' % (
print_location(haplotype), )
h = io_genotype.read('plink', 'haplotype', tped=haplotype,
load_ids=False) if haplotype else None
if verbose:
print 'Reading error data from %s ...' % (print_location(error_file), )
error = np.loadtxt(
error_file) if error_file and os.path.isfile(error_file) else None
if verbose:
print 'Reading frame data from %s ...' % (
print_location(frames_file), )
frames = db_gene.snp.ld_graph.read_frames(
frames_file) if frames_file else None
lam = np.loadtxt(
lam_file) if lam_file and os.path.isfile(lam_file) else None
# info = ProblemInfo(p, g) if info is None else info
problem = Problem(p, g, haplotype=h, error=error, frames=frames, lam=lam)
if haplotype and info:
if verbose:
print 'Reading problem info from %s ...' % (info, )
with open(info, 'rb') as fout:
problem.info = pickle.load(fout)
return problem
